« SMS Disaster Alert System | Main | The Greening of China »

Gravity Tugboat for Dangerous Asteroids

gravtractor.jpgHow do you keep a large asteroid from hitting the Earth? If you're in Hollywood, you blast it with a nuke; sadly, in the real world, that really wouldn't work. A more considered approach would be to attach rockets and push it gently, changing its orbit enough to miss us. But it turns out that the best approach of all could be to park beside it and just sit there.

NASA astronauts Edward T. Lu and Stanley G. Love describe this scenario in the latest Nature. In "A Gravitational Tractor for Towing Asteroids," the scientists argue that a sufficiently large mass, about 20 tons or so, kept within close range of a moderate-sized asteroid, could exert sufficient gravitational influence to pull the object off-course. You would have to keep the gravitational tugboat parked by the asteroid for a year or so. And you'd have to do it 20 years in advance.

With the gravity plan, a spacecraft would not have to dock on the asteroid, but instead hover above its surface. The craft's thruster jets would angle outward to avoid blasting the asteroid's surface and pushing it away.
"You would use this small gravity force between the [spacecraft and the asteroid] as your towline to basically pull this thing," Lu explained. [...]
The astronauts calculate that, given a lead time of about 20 years, NASA could launch a spacecraft that could safely deflect an asteroid some 650 feet (200 meters) across in about a year of "towing."

That small shift, given the lead time, would result in an orbit that would completely miss the Earth by a (comparatively) wide margin.

Lu, who is a member of the B612 Foundation (chaired by fellow astronaut Rusty Schweikart, whom we've discussed before) argues that this would be an ideal approach for asteroid 2004 MN4, now known as 99942 Apophis. That's the object that will narrowly miss the Earth in 2029, coming closer to our planet than the Moon; that close flyby will alter the asteroid's path enough that NASA calculates it will have about a 1-in-5000 chance of hitting us in 2036.

If we can calculate the effect of the flyby enough by early next decade -- and a (comparatively cheap) probe to plant a radio transmitter on the asteroid would greatly help -- we'll have enough to use this gravity tug method to prevent a disaster should we find that we would be hit.

Putting a 20 ton object into space wouldn't be cheap, but it would be far cheaper than cleaning up after an asteroid strike.

(Via Astronomy Picture of the Day)

Comments (15)

Stefan Jones:

This might be a good excuse to do some moon-mining.

If you put up more than one of these tugs it might be cost effective to make fuel and the inert dead-weight parts of your tug out of moon materials.

You could do it remotely. Send up a fleet of little wheeled 'bots to plow dust into a solar-powered smelter. Not fast, not elegant, but eventually you end up with big plugs of aluminum and iron that you'd tack onto your towboat.

Other robots could scrape up frost-rich soil from the south polar region. Use that to make H2 for reaction mass and O2 for all the good things oxygen is good for.

A rock that's 200 meters across doesn't seem like the sort that would cause a dinosaur sized extinction event. It'd probably be disasterous to a city though--my impact physics has long since disintegrated into rust so I don't know the energies involved.

I'm not saying we shouldn't try to divert such rocks, I'm just asking if they've explored what is needed for different sized rocks. Would it take less time if the gravity tugboad were more massive? Would a bigger rock require more time for a given mass?

It's nice to have another tool in the box to deal with these things.

Actually, a 200 meter rock would cause something around 220 megatons of damage, leaving a crater about a mile and a half wide. Even at a hundred kilometers away, it would be louder than a freight train and would feel like a major earthquake.

In short, a 200 meter rock could take out a big city.

(All that according to the University of AZ's impact effects calculator:


Why not use the 20 ton object and ram the asteroid? It seems more effective than using gravity to pull on the object. You wouldn't need a 20 year warning time too.

One more thing I would like to add. It would seem possible to use this asteroid as a weapon. The first to get to the asteroid can direct it to any major city they want. Seems deadly and very possible.

Actually, it would do far less. Most rock asteroids are actually not very dense -- more on the order of a marshmallow than a solid piece of stone -- and would easily absorb the impact with little or no change of direction. Slow pushing/pulling works a lot better than rapid hits of energy.

For a very crude comparison, imagine a car in neutral with the emergency brake off on a flat surface. If you run up to it as fast as you can and slam your body into the back, it will probably roll a bit, but not much, and not very fast. If you instead start pushing and keep pushing, it will eventually start to move, slowly at first, but pick up speed.

That's *not* an exact analogy by any means, but it's suggestive of what the difference is.

As a weapon: possibly, but there are a lot cheaper, easier, and far less detectable ways to unleash a lot of damage.

Stefan Jones:

Nuclear electric, BTW, means an ion motor powered by a nuclear reactor.

Nuclear reactors -- not the same thing as RTGs -- have been orbited before. As I recall, some Soviet spy satellites used them. They were pushed into a higher orbit when depleted, but at least one crashed (northern Canada?).

The papers I've read suggest that space-adapted reactors would be more - or - less solid state, with no circulating fluid. They'd need big radiators. Since the whole point of the mission is to fly a big massy thing into deep space, you could put as much shielding and as many safeguards on the reactor as politics warrant.

In addition to the tug, you'd probably need a couple of service vehicles. They might be tugs without the extra mass. They'd be charged with bringing in more reaction mass (xenon? cesium?) and perhaps replacing reactors fuel elements when called for.

The paper says that a 20 T spacecraft hovering 1/2 radius above a 200 meter asteroid could exert a gravitational tug of 1 N; over a period of 1 year, this could change the velocity of the asteroid sufficiently to avoid a planetary impact 20 years later.

The paper specifies the propellant mass required for rendezvous as 4 tons and 400 kg for the deflection; power budget is 100 kW for a year.  Thing is, you could do the same with an almost trivial spacecraft running on solar power.  It could just bump itself into the asteroid surface and pull up some regolith (it could have long, springy legs to avoid hanging up on things like boulders).  If it gets some regolith, it bounces back a ways and then unloads the regolith at a substantial speed backwards to throw itself back at the asteroid.  Lather, rinse, repeat.

If the spacecraft can repeat the bounce-and-throw cycle 50 times per hour and throws the regolith away at 50 meters per second, it would need to gather 1.44 kg of regolith per bounce to average 1 N of force.  The energy carried away by the regolith is 1250 J/kg * 0.02 kg/sec = 25 watts.

In short, such a spacecraft could be powered by some small solar panels.  Energy storage could be accomplished with batteries or supercapacitors.  Even if it needed to push against the night side of the asteroid, it could perform several bounces and then push itself out into the sunlight for a while to recharge.

(I never thought I'd catch myself saying this, but...) Small is Beautiful.

Pace Arko:

Okay, that impact calculator is boss! Writers of the latest pot-boiler take note because now you have a tool to make your end of the world story slightly more plausible!

Thanks for the link Jamais.

E-P how long would the bouncing 'bot need to do this? One of the advantages of the "just sit there" system is that there's little stress on the tugboat. It seems to me that bouncing up and down nearly once/minute for an extended period of time is going to put some significant material stress on something that's not going to be easy to fix.

But I *do* like the idea...

And Stefan, the reactor would almost certainly be thorium.

Daniel Haran:

My first comment got caught up by the spam filter...

E-P's thinking along similar lines as what I had in mind. If this bot were to mine/drill the asteroid and/or remove regoliths and expel them, it would reduce the mass and normally change the trajectory.

It seems less spectacular and sci-fi than explosions and perhaps a tad less elegant than the gravity tow, but it seems safer to me.

If the bouncing baby 'bot applied the same 1 N average force to the asteroid, it would also take one year to do the job.  More powerful 'bot, less time required.  The proposal would lose 25 watts (average) in kinetic energy; by way of contrast, the Deep Space 1 probe had a solar array of a couple of kilowatts.  If you could apply 1500 watts to the job and scoop stuff up fast enough, you could do it 60 times as fast.  That would cut the job from a year to about 6 days, let you use far more conservative methods, or deal with far bigger rocks.

I don't think we'd have any problems making a mechanism to do this job.  50 bounces an hour is 1200 a day, or 438,000 per year.  There are all kinds of things that cycle more frequently than that and they last many years.  Graphite fiber springs should last long enough.

The beauty of using little one-ton solar powered bouncing 'bots instead of 20-ton nuclear bots is that you could launch a dozen or more of them at a time, and put science instruments on them.  You could deflect a whole heap of rocks and learn one heck of a lot about them for cheap.  (They'd need ion engines to get to their rocks, of course.)


This page contains a single entry from the blog posted on November 11, 2005 1:28 PM.

The previous post in this blog was SMS Disaster Alert System.

The next post in this blog is The Greening of China.

Many more can be found on the main index page or by looking through the archives.

Powered by
Movable Type 3.34